Search research articles

ABOUT JoVE

Overview Leadership Blog JoVE Help Center

AUTHORS

Publishing Process Editorial Board Scope & Policies Peer Review FAQ Submit

LIBRARIANS

Testimonials Subscriptions Access Resources Library Advisory Board FAQ

RESEARCH

JoVE Journal Methods Collections JoVE Encyclopedia of Experiments Archive

EDUCATION

JoVE Core JoVE Business JoVE Science Education JoVE Lab Manual Faculty Resource Center Faculty Site

Terms & Conditions of Use

Search research articles

Related Experiment Videos

Empirical risk minimization for support vector classifiers.

F Perez-Cruz¹, A Navia-Vazquez, A R Figueiras-Vidal

¹Dept. of Signal Theor. and Commun., Univ. Carlos de Madrid, Spain.

IEEE Transactions on Neural Networks

|February 2, 2008

Summary

This summary is machine-generated.

Related Concept Videos

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by

Same author

Electron spin resonance as a powerful tool for studying antioxidants and radicals.

Current medicinal chemistry·2013

Same author

Machine learning and data mining: strategies for hypothesis generation.

Molecular psychiatry·2012

Same author

Generalizing CMAC architecture and training.

IEEE transactions on neural networks·2008

Same author

Cost functions to estimate a posteriori probabilities in multiclass problems.

IEEE transactions on neural networks·2008

Same author

Sample selection via clustering to construct support vector-like classifiers.

IEEE transactions on neural networks·2008

Same author

On the structure of strict sense Bayesian cost functions and its applications.

IEEE transactions on neural networks·2008

Same journal

Universal perceptron and DNA-like learning algorithm for binary neural networks: LSBF and PBF implementations.

IEEE transactions on neural networks·2013

Same journal

Guest editorial: special section on white box nonlinear prediction models.

IEEE transactions on neural networks·2011

Same journal

Data-based fault-tolerant control of high-speed trains with traction/braking notch nonlinearities and actuator failures.

IEEE transactions on neural networks·2011

Same journal

Guest editorial: special section on data-based control, modeling, and optimization.

IEEE transactions on neural networks·2011

Same journal

Neural network-based multiple robot simultaneous localization and mapping.

IEEE transactions on neural networks·2011

Same journal

Data-driven model-free adaptive control for a class of MIMO nonlinear discrete-time systems.

IEEE transactions on neural networks·2011

See all related articles

This study introduces a novel iterative reweighted least squares (IRWLS) method for Support Vector Classifiers (SVCs) with any loss function. This approach yields compact solutions, improving Support Vector Machine performance with empirical risk minimization.

Area of Science:

Machine Learning
Computational Statistics

Background:

Support Vector Classifiers (SVCs) are widely used for classification.
Existing SVC methods often rely on specific loss functions.
Achieving compact solutions (fewer support vectors) is desirable for efficiency.

Purpose of the Study:

To propose a general technique for solving SVCs with arbitrary loss functions.
To enable the empirical risk minimization (ERM) principle for large margin classifiers.
To obtain compact SVC solutions with a reduced number of support vectors.

Main Methods:

Application of an iterative reweighted least squares (IRWLS) procedure.
Formulating SVC solution properties as conditions on the loss function.
Implementing the empirical risk minimization (ERM) inductive principle.

Related Experiment Videos

Main Results:

A general technique for solving SVCs with arbitrary loss functions was developed.
Three key properties of SVC solutions were identified as conditions on the loss function.
The method demonstrated the ability to produce very compact SVC solutions.

Conclusions:

The proposed IRWLS technique offers a flexible approach to SVC training.
This method allows for the optimization of SVCs with diverse loss functions.
The technique effectively balances model complexity and classification performance, as shown with real and synthetic data.